Due to their highly stable nature in blood and relatively few side effects, antibodies have been receiving much attention as pharmaceuticals. Of particular note are bispecific antibodies that can simultaneously recognize two types of antigens, e.g., antibody A and antibody B (see Non-Patent Document 1). MDX-210, which is currently under clinical trial investigation, is an IgG-type bispecific antibody that retargets FcγRI-expressing monocytes and such to HER-2/neu-expressing cancer cells (see Non-Patent Document 2). In general, antibodies are produced using genetic recombination techniques. One specific technique involves the cloning of a DNA encoding an antibody protein from antibody-producing cells, such as hybridomas or sensitized lymphocytes that produce antibodies or a phage library presenting antibody genes, and the insertion of such into a suitable vector, which is then transfected into host cells for antibody production. Production of IgG type bispecific antibodies using genetic recombination techniques involves the introduction of a total of four types of genes into cells, in which these genes of H chains and L chains constitute two types of IgGs of interest, and the secretion of the antibodies by coexpression. In this type of system, expression of the constituent genes of the wild type H chain and L chains leads to random association between two types of H chains and association between H and L chains, and thus, the proportion of the bispecific antibody of interest becomes very small. More particularly, only one out of ten types produced is the bispecific antibody of interest, rendering the production efficiency quite low. Decreased efficiency in the production of the antibody of interest is not only an obstacle for purifying the antibody of interest, but also increases the nonuniformity, such as the lot-to-lot differences, which, in turn, leads to swelling production costs.
Techniques for obtaining L chains commonly shared by both H chains, and the knobs-into-holes technique for heterologous association of H chains have been reported as efficient bispecific antibody production methods for developing bispecific antibodies. More specifically, a common L chain, which can maintain both antigen binding activities of the respective H chains that recognize antigen A and antigen B, is identified from a phage library or such. Then, formation of H chain heterodimers is promoted by substituting an amino acid side chain present in the CH3 region of one of the H chains with a larger side chain (knob), and substituting an amino acid side chain present in the CH3 region of the other H chain with a smaller side chain (hole), to place the knob within the hole. Thus, bispecific antibodies of interest can be efficiently obtained (see Patent Document 1, Non-Patent Document 3, and Non-Patent Document 4).
However, even when the knobs-into-holes technique is used for H chain heterodimers, although the content of the chain A-chain B heterodimer of interest can be increased up to a maximum of about 95% as shown in Non-Patent Document 3 and Non-Patent Document 4, the remaining 5% is impurities and consists of chain A and chain B homodimers. To develop bispecific antibodies as pharmaceuticals, the chain A-chain B heterodimers must be purified to the highest possible purity from the three types of molecular species (chain A homodimer, chain B homodimer, and chain A-chain B heterodimer) that are produced when a common L chain is used (Non-Patent Document 3 and Non-Patent Document 4). Therefore, it is necessary to remove the 5% impurities, which are the chain A and chain B homodimers, thereby purifying the chain A-chain B heterodimer to a high purity that allows the heterodimer to be developed into a pharmaceutical. When a common L chain is used without the knobs-into-holes technique, the production ratio of the chain A homodimer, chain A-chain B heterodimer, and chain B homodimer is theoretically 1:2:1, and the 50% impurities which are the chain A and chain B homodimers must be removed.
Several chromatographic methods for separating the chain A-chain B heterodimer from the chain A and chain B homodimers at the level of pharmaceutical manufacturing have been reported. Non-Patent Document 5 reports a method for selectively purifying the chain A-chain B heterodimer using mouse IgG2a as chain A and rat IgG2b as chain B. This method uses difference between the respective mouse IgG2a and rat IgG2b H chains in their affinity for protein A, and purifies the chain A-chain B heterodimer by controlling the pH for elution from protein A. However, since constant regions from mouse and rat are used, this method is difficult to apply to pharmaceuticals for human from the perspective of antigenicity. Furthermore, since this method cannot separate the chain A-chain B heterodimer, which is composed from H chains belonging to the same subclass, its use is limited.
A method for purifying the chain A-chain B heterodimer using hydrophobic interaction chromatography is reported in Non-Patent Document 6. However, the peak of the chain A-chain B heterodimer of interest containing anti-CD3 mouse IgG2a and anti-CD19 mouse IgG1 is not sufficiently separated. In addition, H chains belonging to different subclasses are used, and the difference in their hydrophobicity seems to be used for the separation. Thus, this method may not necessarily separate the chain A-chain B heterodimer composed from H chains belonging to the same subclass.
A method for purifying the chain A-chain B heterodimer using thiophilic affinity chromatography is reported in Non-Patent Document 7. However, this method cannot be adopted to separate the chain A-chain B heterodimer composed from H chains belonging to the same subclass, because it uses mouse IgG1 and rat IgG2a, and the free cysteines (thiol groups) in the hinge regions. In addition, since the free cysteines are involved in aggregation during storage, this method is not suitable for development of stable pharmaceutical formulations.
Affinity chromatography using antigens is reported in Non-Patent Document 8. However, since affinity chromatography using proteins or peptide antigens is problematic in terms of cost and column stability, production of pharmaceuticals using affinity chromatography is unconventional. Furthermore, to purify the chain A-chain B heterodimer that binds to both antigens, affinity chromatography must be performed twice, and this is expected to become costly. It has been reported that there are antibodies that recognize only the three-dimensional structures of antigens as well as antibodies that have desired functions but low affinity. For antibodies with such characteristics, it is difficult to adopt affinity chromatography that uses antigens. Therefore, purification of bispecific antibodies using affinity chromatography cannot be used widely.
As described above, purification of the chain A-chain B heterodimer of a bispecific antibody has been performed only within limited scope. There has been no report on methods for purifying the chain A-chain B heterodimer of a bispecific antibody composed from the same H chain subclass and constant region sequence to a high purity that is acceptable for pharmaceuticals. When two types of antibodies constituting a bispecific antibody have the same constant region sequence, the chain A-chain B heterodimer needs to be separated based solely on the differences in their variable region sequences. However, since the amino acid sequence homology between antibody variable regions is very high (Non-Patent Document 9), it has been difficult to purify the chain A-chain B heterodimer to a high purity that is acceptable for pharmaceuticals solely based on the differences in their variable region sequences.
[Patent Document I]
WO 96/27011
[Non-Patent Document 1]
Marvin J S, and Zhu Z, “Recombinant approaches to IgG-like bispecific antibodies.”, Acta. Pharmacol. Sin., June 2005, Vol. 26(6), p. 649-58.
[Non-Patent Document 2]
Segal D. M. et al., Current Opinion in Immunology, 1999, Vol. 11, p. 558-562.
[Non-Patent Document 3]
Merchant A M et al., “An efficient route to human bispecific IgG”, Nat. Biotechnol., July 1998, Vol. 16(7), p. 677-81.
[Non-Patent Document 4]
Carter P, “Bispecific human IgG by design.”, J. Immunol. Methods., February 2001, Vol. 248(1-2), p. 7-15.
[Non-Patent Document 5]
Lindhofer H et al., “Preferential species-restricted heavy/light chain pairing in rat/mouse quadromas. Implications for a single-step purification of bispecific antibodies.”, J. Immunol., Jul. 1, 1995, Vol. 155(1), p. 219-25.
[Non-Patent Document 6]
Manzke O et al., “Single-step purification of bispecific monoclonal antibodies for immunotherapeutic use by hydrophobic interaction chromatography.”, J. Immunol. Methods., Oct. 13, 1997, Vol. 208(1), p. 65-73.
[Non-Patent Document 7]
Kreutz F T et al., “Efficient bispecific monoclonal antibody purification using gradient thiophilic affinity chromatography.”, J. Chromatogr. B. Biomed. Sci. Appl., Sep. 4, 1998, Vol. 714(2), p. 161-70.
[Non-Patent Document 8]
Gupta S and Suresh M, “Affinity chromatography and co-chromatography of bispecific monoclonal antibody immunoconjugates.”, J. Biochem. Biophys. Methods., May 31, 2002, Vol. 51(3), p. 203-16. Review.
[Non-Patent Document 9]
Carl Branden, Introduction to Protein Structure 2nd edition, Newton Press.